The international maritime organizations (IMO) updated non-mandatory guidelines for reduction of underwater radiated noise (URN) from a ship in 2023. The updated guidelines accepted a new idea, a URN management plan. If URN is monitored using fixed-point underwater sound measurement, the measured underwater sound should be classified. Hence, in this study, we investigated a method to automatically classify underwater sounds measured under a shipping lane using machine learning (k-means clustering), and the results were validated using AIS data. The underwater sounds were recorded under the shipping lane south of Izu-Oshima Island, Japan.
The International Marine Organisation (IMO) updated the nonmandatory guidelines for reducing URN from a ship in 2023. If URN is monitored using fixed point underwater sound measurement, the measured underwater sound should be classified. This is because there are several sources of underwater sounds other than ships, such as wind and ocean life. Also, the effect of flow noise should be omitted.
In a previous study, Li et al. (2020); Luo et al. (2021) classified the size of ships using short-duration underwater sound. Lin et al. (2021) classified underwater sound sources using also short-duration underwater sound.
In this study, we investigated the automatic classification of underwater sound sources using underwater sound measured under a shipping lane using machine learning (k-means clustering), and the results were validated using AIS data. Underwater sounds were recorded under the shipping lane south of Izu-Oshima Island, Japan (Sakai et al., 2023). These may improve safety by providing navigational information for vessels that do not transmit AIS data, and may also contribute to the detection of suspicious vessels.
Measurements of underwater sound were conducted in the water south of Izu-Oshima Island, Japan (Fig. 1) (Sakai et al., 2023). Table 1 shows the measurement period, the sinker points for the monitoring station with hydrophones, and the water depth at the sinker point. The position of the sinker was estimated by measuring the slant range from a work vessel to the disconnection device at several points using sound waves. Three hydrophones were deployed with two pressure gauges (depth meter) and a current meter (Table 2). The hydrophones were deployed using a sinker and buoyant materials to form the monitoring station as vertically as possible from the sea floor (Fig. 2).
